--- Day changed Sat Oct 15 2016
00:08 -!- JM-IRCRelay [~RelayBot@unaffiliated/belcher] has quit [Ping timeout: 260 seconds]     
00:09 -!- JM-IRCRelay [~RelayBot@unaffiliated/belcher] has joined #joinmarket     
00:11 -!- Cory [~Cory@unaffiliated/cory] has quit [Ping timeout: 260 seconds]     
01:25 -!- AdrianG_ [~User@45.ip-158-69-201.net] has joined #joinmarket     
01:29 -!- viasil_ [~viasil@198.12.75.25] has joined #joinmarket     
01:33 -!- Netsplit *.net <-> *.split quits: solace, viasil, deafboy, megaddin, AdrianG, berndj
01:44 -!- Netsplit over, joins: solace
01:51 -!- berndj-blackout [~berndj@mail.azna.co.za] has joined #joinmarket     
01:51 -!- megaddin [aladdin@gateway/shell/fnordserver.eu/x-ealzpiqjlfyoteea] has joined #joinmarket     
01:51 -!- deafboy [quasselcor@cicolina.org] has joined #joinmarket     
02:32 -!- berndj [~berndj@mail.azna.co.za] has joined #joinmarket     
02:32 -!- deafboy_ [quasselcor@cicolina.org] has joined #joinmarket     
02:37 -!- Netsplit *.net <-> *.split quits: deafboy, berndj-blackout, megaddin
02:50 -!- rdymac_ [uid31665@gateway/web/irccloud.com/x-mhrpoyjepophhhuy] has joined #joinmarket     
02:55 -!- rdymac_ [uid31665@gateway/web/irccloud.com/x-mhrpoyjepophhhuy] has quit []     
02:55 -!- rdymac_ [uid31665@gateway/web/irccloud.com/x-fmdjzzpwwlfnczhn] has joined #joinmarket     
03:38 -!- megaddin [aladdin@gateway/shell/fnordserver.eu/x-nxfevqjvgbscfvkh] has joined #joinmarket     
03:38 < JM-IRCRelay> [AlexCato] belcher, you're of course right. Never even noticed ob-watcher's existence until now
04:55 < waxwing> adlai: i updated the gist to reflect that important correction, see the last bullet point in the "Proposed alteration.." subheading. It's a pretty tricky point, I only took a brief look so I'm not sure how hard the code alterations are, but in principle it should be OK I guess.
04:56 < waxwing> of course we could just fall back to doing it the same way as now, where the only thing that happens is that bad actors are added to ignored_makers and then start again from the beginning, but that loses the property of being able to avoid using a new commitment.
04:56 < waxwing> (sorry prob stating the obvious there...)
04:57 < JM-IRCRelay> [waxwing] AlexCato: yeah i always try to mention the ob-watcher thing, it's understandable people don't want to bother. i think we could optionally remove the matplotlib dependency so it's even less of a bother to do it locally.
04:58 < JM-IRCRelay> [belcher] perhaps we could make it that the script will still run without matplotlib but wont generate graphs
04:58 < JM-IRCRelay> [waxwing] yeah that's what i was thinking.
05:03 < waxwing> belcher: it's a bit of a memory stretch, but offhand can you see a breaking problem with re-issuing a new !tx to the same maker without going back to !fill ? (assuming amount hasn't changed ofc)
05:04 < waxwing> it goes on_seen_tx to recv_tx in CoinJoinorder, then it checks the verify_unsigned
05:04 < waxwing> i did wonder a bit about adding the unconfirmed watching callback
05:04 < waxwing> i mean add_tx_notify
05:05 < belcher> i think the makers record where they're at in the protocol
05:06 < waxwing> on_seen_tx only checks whether there's an active order currently
05:06 < belcher> i guess its ok then
05:06 < waxwing> ah no i'm wrong; it sets the order to None at the end of recv_tx :(
05:08 < waxwing> hmm that might make the whole idea unworkable, shame
05:09 < waxwing> well i overstate it; that's not the *whole* idea; the main idea is to gather M > N potential counterparties and read their utxos and ignore those with overlaps
05:09 < waxwing> the idea of using some of the excess M-N to complete the tx if there's a disruption was secondary
05:24 < waxwing> what does !rejected mean here: 
05:24 < waxwing> ! [rejected]        refs/pull/631/head -> pull_631  (non-fast-forward)
05:25 < belcher> you need to do git pull
05:25 < belcher> what happened is the main branch on github moved forward and your new commit wasnt added to the tip
05:25 < belcher> it might cause a conflict now that you've already made the commit
05:26 < waxwing> i was doing this: git fetch jmorg refs/pull/631/head:pull_631
05:26 < waxwing> (should have said that really)
05:26 < belcher> what is git branch ?
05:26 < belcher> maybe switch to another branch
05:26 < waxwing> branch currently is develop, i'm creating a new branch to test the PR
05:26 < belcher> use the same name as what jmorg calledit
05:27 < waxwing> used this command a few times before for this
05:27 < belcher> hmm not sure
05:27 < waxwing> currently #631 says "no conflicts with the base branch"
05:33 < JM-IRCRelay> [AlexCato] i failed at git with #631, so whatever state thats caused, i'm sorry. Can checkout a fresh jm-develop and create a new PR with just one commit if you like
05:34 < JM-IRCRelay> [waxwing] AlexCato: it seems to be OK now; i think it's because i hadn't fetched the whole repo, when i did fetch jmorg and then retried, i no longer got the 'rejected' message and all seems correct
05:34 < JM-IRCRelay> [AlexCato] great
05:35 < JM-IRCRelay> [waxwing] yes you have 1 more commit than needed, but it isn't worth the bother imo to change it now.
05:35 < JM-IRCRelay> [AlexCato] adding the log filterting mechanism somehow added more commits at the top as well (there used to be 1 after my initial PR post, but now there's 4...)
05:35 < JM-IRCRelay> [waxwing] at some point we can try to start doing git better, using rebase or in any case not using the default merge process we use now which everyone tells me is wrong, i sort of half understand.
05:36 < JM-IRCRelay> [waxwing] i have to make various incantations to get the merge commit signed :)
05:44 < waxwing> first impression is that the pubmsg and privmsg printouts on the terminal are still too noisy
05:45 < waxwing> the overall arrangement seems right though; the logfile still has the extra stuff like podle and unspent printouts
05:46 < waxwing> need the broadcast tx txid to show up on INFO too
05:48 < waxwing> i'm gonna try it with privmsgs at DEBUG (since they're mostly all encrypted b64 anyway), and txid bumped to INFO. That'll probably look a fair bit better.
05:58 < waxwing> heh, one full sendpayment all one one (admittedly large) screen :)
05:58 < waxwing> *on one
06:11 < JM-IRCRelay> [AlexCato] feel free to modify anything in that PR.  If it's better without the privmsgs, go ahead. Only need to keep in mind that the user of the scripts need some feedback that it's still running and doing something
06:16 < waxwing> heh, "CRASHING, DUMPING EVERYTHING" might not be an ideal message :) maybe "Quitting, dumping object contents to log".
06:17 < JM-IRCRelay> [AlexCato] yeah, thats better
06:19 < waxwing> running test_tumbler now, it's looking much better. funnily enough the only thing left i'd class as "terminal spam" is the !hp2 pubmsgs.
06:21 -!- Yohkii [~IceChat9@unaffiliated/yohkii] has quit [Ping timeout: 260 seconds]     
06:30 < waxwing> if you upgrade Core but don't change the blockchain data, would you need to rescan?
06:32 < belcher> shouldnt need to
06:32 < belcher> btw if anyone knows bitcoind and wants a job https://www.reddit.com/r/Bitcoin/comments/5743u2/blockchaininfo_is_down_again/d8pjji0
06:36 < waxwing> yeah it looks like he's actually on a new ubuntu install (see #633) ... i wonder if he synced from scratch (in which case clear) or just moved his blockchain over or something. oh well. likely it's rescan.
06:38 < owowo> it's written out there, b.info is been DNS-hijacked
07:05 < GithubBot5678> [joinmarket] AdamISZ closed pull request #631: Add different log levels (develop...develop) https://git.io/vPB3k
07:06 < waxwing> i'll add the few little tweaks mentioned above in a bit
07:28 -!- Giszmo [~leo@pc-40-227-45-190.cm.vtr.net] has joined #joinmarket     
07:56 -!- kakobreklaa is now known as kakobrekla     
07:56 -!- kakobrekla [~kako@163.172.132.224] has quit [Changing host]     
07:56 -!- kakobrekla [~kako@unaffiliated/kakobrekla] has joined #joinmarket     
07:58 < belcher> well done on the log levels, glad it was finally done
07:58 < belcher> how would you change the log levels? in joinmarket.cfg ?
08:04 < belcher> wonder how this could've happened https://www.reddit.com/r/joinmarket/comments/57bsqw/error_not_enough_liquidity_in_the_orderbook/ did all his UTXOs get banned?
09:05 < JM-IRCRelay> [AlexCato] [belcher] how would you change the log levels? in joinmarket.cfg ? <-- yes, i added it there
09:07 < JM-IRCRelay> [AlexCato] about his 24 hrs 'not enough liquidity' --> my YG had loads of commitment failures, might have been him. Exact message was, that the commitment was too small (lower than 20% of cj amount). Maybe that was him. The error sent back to the taker seems to be a generic one though
09:09 < JM-IRCRelay> [belcher] wouldnt his script check the 20% figure ?
10:06 < waxwing> belcher: on reflection this is really an error on my part - here's the situation:
10:06 < waxwing> in tumbler, the most likely way to not source a commitment is because it's not old enough. but ofc that's not the only possibility.
10:07 < waxwing> so i coded it to wait a few minutes and try again. it's also sensible to wait in a scenario where a user decides to add commitments to the external set, that way they don't have to stop the tumbler. but that's not very convenient to say the least.
10:08 < waxwing> probably i should have written it to just give up if it can't source and specifically the reason it can't source a commitment is *not* that it's too young.
10:08 < belcher> he says he left the script for an hour
10:08 < waxwing> well, not probably, definitely.
10:08 < belcher> ah     
10:08 < waxwing> yeah so currently an infinite loop scenario is very possible, specifically if a number of failures are occurring.
10:09 < waxwing> i guess one POV i had was, well if it's stuck for that reason, nothing good is going to happen anyway, so just keep trying until the user comes back, sees that it's failed, and restarts.
10:10 < waxwing> but currently in that situation, they don't know why it's failed. i'll double check whether it writes out to commitments_debug.txt, in which case they can just read the file to see the current issue/problem.
10:10 < waxwing> if they see it's too small amount for example, they know that.
10:13 < waxwing> yes it does write to file, he can at least read that and check whether it's the size thing
10:17 < waxwing> so it's tricky, but maybe the only error is not making that information more directly visible.
10:21 -!- grubles [~grubles@unaffiliated/grubles] has quit [Quit: Leaving]     
10:25 -!- grubles [~grubles@unaffiliated/grubles] has joined #joinmarket     
10:26 < GithubBot5678> [joinmarket] AdamISZ pushed 1 new commit to develop: https://git.io/vP6oV
10:26 < GithubBot5678> joinmarket/develop e1ca488 Adam Gibson: move privmsgs to debug level, minor tweaks to a few log msgs. Fix bug in tumbler fast sync option access
10:28 < JM-IRCRelay> [waxwing] AlexCato: i take your point about users need to know what's going on, but i think that'd have to be addressed with more intelligent logging; i think showing all the encrypted b64 is way too noisy. i think the way it works now, in defaults, is fine if a little basic; and the log files have it all ofc.
10:32 < waxwing> belcher: i'd always thought that < 20% would be ultra-rare in using tumbler, but i guess it could become likely if you specified e.g. 6 or more txs/mixdepth. i doubt many people do that though.
10:39 < waxwing> what are the build dependencies for libsodium, anyone know?
10:41 -!- lnostdal [~lnostdal@90.149.73.62] has quit [Quit: lnostdal]     
10:42 < waxwing> i get the impression that people are not appreciating the subtlety: "sourcing commitments fail" does not mean *using* commitments; thus alexcato's comment about seeing a lot of commitment failures would not apply in that scenario; if the Taker can't find an appropriate one, he won't send anything.
10:43 < waxwing> ah, but yes it can be indirect evidence; if he had tried a lot earlier and they had failed then he would have run out.
10:44 < waxwing> hopefully he didn't edit the config :)
11:17 < waxwing> adlai: ah yeah, i forgot about your filtering ideas re: choosing extra makers, well feel free to either make a new gist or i can add text to what i've already written
11:22 < waxwing> currently 30 offers from *same* bot with max size 159.5. would be nice to let him know that that doesn't bump his chance of getting chosen.
11:22 < belcher> lol     
11:54 < owowo> yep, sometimes more can be too much ;)
12:09 -!- Yohkii [~IceChat9@unaffiliated/yohkii] has joined #joinmarket     
12:11 -!- Yohkii [~IceChat9@unaffiliated/yohkii] has quit [Client Quit]     
15:13 -!- coins123 [~coins123@unaffiliated/coins123] has joined #joinmarket     
15:13 -!- coins123 [~coins123@unaffiliated/coins123] has quit [Remote host closed the connection]     
15:15 < JM-IRCRelay> [AlexCato] waxwing: i think libsodium only depends on build-essentials. I've linked to an installation guide for libsodium here: https://www.reddit.com/r/joinmarket/comments/57cbxg/would_really_like_to_use_joinmarket_but_keep/d8szhes
15:22 -!- TailsUser [1810d321@gateway/web/freenode/ip.24.16.211.33] has joined #joinmarket     
15:26 < TailsUser> Has any patient person successfully installed the latest version of joinmarket on Tails? In the past I've always used the tailsjoin tool, but it's not compatible with >2.0, so I'm trying without it... and failing :(
15:27 < TailsUser> I'm on the pip install -r requirements.txt part. I did the previous steps. I get an error "Command python setup.py egg-Info failed with error code 1". I'm opening up the debug log for more info now.
15:29 < TailsUser> InstallationError: Directory /home/amnesia/.pip/ is not installable. File setup.py not found.
15:30 < JM-IRCRelay> [AlexCato] add 'sudo' in front of the 'pip install'
15:30 < JM-IRCRelay> [AlexCato] hope you did choose to set up an admin password when you booted up tails, cause otherwise you'll have to reboot, do that, and begin again
15:31 < TailsUser> Yeah I did. It prompted me for a sudo password on the original call, so I assumed it was elevated.
15:32 < TailsUser> It failed again, except the log in is /root/.pip/pip.log now
15:32 < JM-IRCRelay> [AlexCato] sadly, i didnt set up the admin pwd on my own tails. Will reboot and do that now and see if i can get it working. Brb
15:34 < TailsUser> It looks like libffi is not being found... the log file is pretty damn big
15:34 < TailsUser> Then the final issue is Python.h No such file or directory on trying to compile _cffi_backend.c
15:36 < TailsUser> Reddit shows someone posted a installing secp256k1 on tails page on the wiki, but the link no longer works.
15:36 < TailsUser> https://www.reddit.com/r/joinmarket/comments/4j1ugu/installing_secp256k1_on_tails/
15:37 -!- coins123 [~coins123@unaffiliated/coins123] has joined #joinmarket     
15:37 < JM-IRCRelay> [AlexCato] so, on ii
15:38 < JM-IRCRelay> [AlexCato] *it
15:38 < TailsUser> Hey Alex, I found some specific install instructions for libsecp256k-1 on tails
15:38 < TailsUser> There are a few magic commands I didn't try... trying them now
15:38 < TailsUser> https://github.com/JoinMarket-Org/joinmarket/wiki/Installing-the-libsecp256k1-binding
15:42 < TailsUser> nice, this install is getting farther than the requirements.txt one
15:44 < TailsUser> Okay that's setup. Now I just need to get libnacl
15:46 < TailsUser> hmm. I did a sudo torify pip install libnacl ... but wallet-tool.py can't find the module. Maybe I need to install to a specific location?
15:47 < TailsUser> running the pip install -r requirements.txt tells me everything is installed.
15:49 < TailsUser> strangeness. If I type python then import libnacl, it fails. I'm way outta my element here. I'll keep googling.
15:50 < JM-IRCRelay> [AlexCate] so, an 'apt-get upgrade' killed my tails' tor connection. If you wrote anything in the meantime, i didnt see it. Going to write a tails installation instruction if we get it working
15:51 < TailsUser> I spammed a lot
15:51 < TailsUser> Screencap of conversation: http://screencast.com/t/omf9V4FM4f
15:52 < JM-IRCRelay> [AlexCate] d'oh. There are no public logs of this channel anywhere i guess, belcher?
15:52 < JM-IRCRelay> [AlexCate] or like that. Thanks
15:52 < TailsUser> Basically, one of the packages has documentation. I got libsecp256k1 working
15:53 < TailsUser> but even with libnacl installed on the OS, when I try to import libnacl in python, it fails.
15:53 < TailsUser> maybe it's installing for the wrong python version
15:54 < TailsUser> screw pip, I'm going to download the tar
15:57 < TailsUser> nope. grr
15:58 < JM-IRCRelay> [AlexCate] still in libsec, doing this stuff over tor is kinda slow...
15:59 < JM-IRCRelay> [AlexCate] maybe the 2nd link in here helps with libnacl? https://askubuntu.com/questions/330589/how-to-compile-and-install-dnscrypt/330611#330611
15:59 < TailsUser> I got it working, I think
16:00 < JM-IRCRelay> [AlexCate] what did the trick?
16:00 < TailsUser> Okay, so you use the instructions on the wiki to manually install the libsecp256k1
16:01 < TailsUser> Then you download libnacl tar file. Install it somewhere. Then copy the libnacl folder out of that directory and into the joinmarket directory
16:02 < TailsUser> wallet-tool.py is working now. I'm going to send it $10 and I guess have it send it back to me to validate sendpayment.py
16:02 < JM-IRCRelay> [AlexCate] awesome. Thanks for letting us know, i'll create/update a tails 'quick install' guide so that all this is in one place in the future
16:04 < TailsUser> For sure. I'm happy to be back using secure software. It's been 6 months since I've updated any of this.
16:05 < TailsUser> I'll let you know in 5-10 mins when the transactions get confirmed to see if sendpayment is working
16:05 < TailsUser> I haven't done any config edits, so I might still have issues
16:10 < JM-IRCRelay> [AlexCate] got it working too. No need to get the libnacl tar even, can just copy the existing libnacl from tails after the pip install
16:10 < TailsUser> Sounds right. I didn't know where to find it
16:11 < JM-IRCRelay> [AlexCate] neither did I... but some linux file searching showed it to be in /lib/live/mount/overlay/usr/local/lib/python2.7/dist-packages/libnacl
16:12 < JM-IRCRelay> [AlexCate] anyways, still ugly and cumbersome to get this all set up in tails. Possibly should add some tails_install.sh script for that - opinions about that, waxwing/adlai/belcher ?
16:13 < TailsUser> That would be nice. I'd also recommend pulling down tailsjoin from the github since it's worthless now.
16:28 < TailsUser> giving me 'could not find orders to complete transction.'
16:29 < JM-IRCRelay> [AlexCato] sorry, just rebooted again. Only saw your last line. When do you see that error?
16:29 < TailsUser> I tried a sendpayment.py -N 5
16:29 < TailsUser> I'm trying 3 now
16:29 < JM-IRCRelay> [AlexCato] wait a sec
16:29 < TailsUser> This is for about $1, so I doubt there's a liquidity problem
16:29 < JM-IRCRelay> [AlexCato] how many confirmations do you have yet?
16:30 < TailsUser> 2     
16:30 < TailsUser> It's working this time.
16:30 < JM-IRCRelay> [AlexCato] wait one more. Standard config of the new 0.2.x versions will not accept UTXOs which are not at least 3 confirmations old
16:30 < TailsUser> ahh. Okay.
16:31 < TailsUser> That's like a lifetime these days ;)
16:31 < JM-IRCRelay> [AlexCato] buuut... it's working? Maybe i misunderstood something there, but am pretty sure
16:31 < JM-IRCRelay> [AlexCato] yeah, its one of the mechanisms to counter the snooping attacks
16:32 < TailsUser> We'll find out. I just tried again and it gave me a TypeError: cannot concatenate str and NoneType objects
16:33 < JM-IRCRelay> [AlexCato] cant recall what caused that, but i think this was fixed in the development branch already. Sec.
16:34 < TailsUser> I still only have 2 confirmations, so I'm waiting before I try again
16:36 < JM-IRCRelay> [AlexCato] yeah, might be that. Cant find the error message atm. Also would recommend to change your tx_fees setting in the joinmarket.cfg to 25. Currently there no TX queue, so this could save you some transactions fees while still confirming in the next block
16:40 < JM-IRCRelay> [AlexCato] so, there's the 3rd confirmation
16:40 < TailsUser> k. trying again
16:40 < TailsUser> I changed my tx_fees to 25
16:42 < TailsUser> same error
16:42 < TailsUser> I choose Y then get the error
16:43 < TailsUser> Cannot construct transaction, failed to generate commitment, shutting down. 
16:45 < TailsUser> Exception in payment thread. self.create_txt, self.finishcallbck, taker.py -> choose_orders_recover, self.kp.hex_pk(), message_channel.py -> fill_orders -> line 803 -> msg += ' ' + commitment
16:45 < JM-IRCRelay> [AlexCato] do you have a file called commitments_debug.txt in your joinmarket directory or logs-folder?
16:45 < TailsUser> I'm guessing commitment didn't get initialized correctly. So it's a nonetype instead of a str.
16:45 < TailsUser> ya I have it
16:46 < JM-IRCRelay> [AlexCato] that might tell us something
16:46 < TailsUser> okay one minute
16:47 < JM-IRCRelay> [AlexCato] though I cant recall if the 2.1 release has / had this kind of bug; seems like it shouldnt work for anybody if a string conversion was missed here
16:48 < TailsUser> http://pastebin.com/jmdBU9bX
16:51 < JM-IRCRelay> [AlexCato] seems i was wrong about the 3 confirmations. It's 5
16:51 < JM-IRCRelay> [AlexCato] you're at 4 currently if i counted right
16:52 < TailsUser> ahh. Wow, that's a lot
16:52 < JM-IRCRelay> [AlexCato] if it doesnt work again with a 5th conf, i'd recommend checking out the development version of joinmarket and see if its fixed there. Just make sure you dont accidently delete your wallet file with the old version
16:53 < TailsUser> I liked to just do coinbase -> node1 -> wallet-gen -> tumble -N 5 -> DONE ... and do that all at once
16:53 < TailsUser> Then come back the next morning and it would have confirmed the whole way through
16:53 < JM-IRCRelay> [AlexCato] im not too familiar with the tumbler, belcher might be of help there. This still could work, if the tumbler is smart enough to start at 5 confirmations
16:57 < GithubBot5678> [joinmarket] AlexCato opened pull request #635: Add quick install notes for tails (develop...patch-1) https://git.io/vP6bK
16:59 < JM-IRCRelay> [AlexCato] 5th confirmation should be in now
17:06 < TailsUser> Trying now.
17:08 < TailsUser> fee shot up to 0.1% ;)
17:09 < JM-IRCRelay> [AlexCato] there's some chance involved, unless you have chosen the -C option :)
17:09 < TailsUser> What's -C ?
17:10 < JM-IRCRelay> [AlexCato] always uses the cheapest offers
17:10 < JM-IRCRelay> [AlexCato] shouldnt be used though by default, because otherwise someone could just set up 6 yield generators to be the cheapest, and then get all traffic = can de-anonymize everyone
17:11 < JM-IRCRelay> [AlexCato] with likelihood weighting, this is a lot harder to pull off
17:14 < TailsUser> I've never had a problem with the fee.
17:14 < TailsUser> So I'm on the 3rd attempt now
17:15 < TailsUser> it keeps going for awhile then asking me a new price cause someone backed out
17:15 < TailsUser> And i got the str-nonetype error again
17:15 < TailsUser> sigh
17:15 < JM-IRCRelay> [AlexCato] yeah, the coinjoin fees are pretty cheap for most amounts
17:15 < JM-IRCRelay> [AlexCato] alright, one more thing to try:
17:16 < JM-IRCRelay> [AlexCato] can you go one directory up in your terminal, so you're not in joinmarket-0.2.1 any more
17:16 < JM-IRCRelay> [AlexCato] then use these two:
17:16 < JM-IRCRelay> [AlexCato] git clone https://github.com/JoinMarket-Org/joinmarket.git
17:16 < JM-IRCRelay> [AlexCato] cd joinmarket
17:16 < JM-IRCRelay> [AlexCato] git checkout remotes/origin/develop
17:17 < JM-IRCRelay> [AlexCato] the copy your wallet-file (probably wallet.json) from your joinmarket-0.2.1 to the new joinmarket directory (subfolder wallets/ as well)
17:17 < JM-IRCRelay> [AlexCato] and do the same thing for the joinmarket.cfg (from joinmarket-0.2.1 to joinmarket)
17:18 < JM-IRCRelay> [AlexCato] the try the sendpayment again. If its still giving you the same error, this seems like a programming bug to be fixed
17:27 < TailsUser> Nope, didn't work
17:27 < TailsUser> I get a cannot construct transaction, failed to generate commitment
17:27 < TailsUser> I see a lot of ERROR counterparty requires sub-dust change.
17:28 < TailsUser> I'm about to give up and try again another day
17:28 < JM-IRCRelay> [AlexCato] thats a problem with their configuration then though, not yours. Can you check the commitments_debug.txt file now?
17:28 < TailsUser> sure
17:29 < TailsUser> http://pastebin.com/wyBdTgpX
17:31 < JM-IRCRelay> [AlexCato] thanks. Nothing in there I can see which seems wrong. I've no more ideas to try out; maybe waxwing/belcher can comment later if they have any ideas
17:32 < JM-IRCRelay> [AlexCato] i'll play around a bit with my own transactions to see if i can replicate the problem
17:32 < JM-IRCRelay> [AlexCato] thanks for helping with the tails setup!
17:34 < TailsUser> For sure. Thanks for your help. I'll be around tomorrow.
19:13 < JM-IRCRelay> [AlexCato] aftermath of the discussion about using TAILS above:
19:15 < JM-IRCRelay> [AlexCato] - theres some makers which add 1000 tx fee to joins, but allow that even for very low amounts, making the effective CJ fee negative (the sybil ones with those strange 7.x BTC max offers). They just time out to join requests then tough, making any sendpayment/tumbler-run tedious, because they will be chosen in the vast majority of joins
19:17 < JM-IRCRelay> [AlexCato] - i've successfully used TAILS myself to do a sendpayment. The problem of TailsUser above was self-made: first the UTXOs were not old enough. Once they were, they had been used too often already (even though they're not old enough they still get used up, it seems?)
19:17 < JM-IRCRelay> [AlexCato] not sure about the last thing. Another theory might be that there's just not enough liquidity for joins that small ($1 value)
20:57 -!- Giszmo [~leo@pc-40-227-45-190.cm.vtr.net] has quit [Quit: Leaving.]     
21:14 -!- grubles [~grubles@unaffiliated/grubles] has quit [Ping timeout: 244 seconds]